Solar Energy 251 (2023) 420–437

Contents lists available at ScienceDirect

Solar Energy
journal homepage: www.elsevier.com/locate/solener

Lighting and visual comfort performance of commercially available tubular

daylight devices
Luís L. Fernandes, Cynthia M. Regnier
Lawrence Berkeley National Laboratory, 1 Cyclotron Rd, MS 90R3147, Berkeley, CA 94720, USA

A R T I C L E I N F O A B S T R A C T

Keywords: Tubular daylight devices (TDDs) require a much smaller roof opening than conventional skylights and, because
Daylighting of their highly reflective tube, they can deliver daylight farther away from the building envelope. This can
Tubular daylight devices provide lighting energy savings, increasing resilience in new and existing buildings. Different types and con­
Light pipes
figurations are available amongst commercially available TDDs, including domes/diffusers with varying optical
Tubular skylights
Experimental study
properties and the diameter of the TDD. This paper presents a comprehensive experimental evaluation of the
Visual comfort lighting and visual comfort performance of multiple configurations of commercially available TDDs, varying
dome type (prismatic and clear), diffuser type (Fresnel and prismatic), and diameter (53 and 35 cm), under a
range of environmental conditions (different times of day/year, sky cover). Based on illuminance measurements,
estimated lighting energy use is also presented. Results indicate that, for clear sky, light levels increase and
energy use decreases with solar altitude (e.g., 16 Wh/m2 daily energy use intensity on a high maximum daily
solar altitude (MDSA) day and 34 Wh/m2 on a low MDSA day) and TDD diameter (e.g., 34 Wh/m2 and 69 Wh/
m2 for 53 cm and 35 cm TDDs, respectively, for low MDSA). The daily illuminance profile is more rounded for
prismatic domes and has higher peaks for clear domes; this translates into a somewhat higher average daily
useful daylight illuminance (DUDI) for prismatic domes (86 %) when compared to clear domes (80 %). No clear
impact of diffuser type was apparent. Measurements indicated no discomfort glare for any of the conditions
tested.

1. Introduction the appearance and light diffusing characteristics of electric lighting

1.1. Background
The use of daylight to offset energy use with electric lighting is one of
the available approaches for helping to reduce carbon emissions caused
by energy use in buildings. Core sunlighting comprises several tech­
niques for bringing daylight – and particularly direct sunlight – deeper
into buildings than what is achievable with conventional openings like
windows and skylights, and in a way that makes use of this light while
minimizing glare. Sunlight can be collected on the roof or on exterior
walls, sometimes undergoing some level of concentration. The collected
light can then be transported via highly efficient optical fibers or ducts/
guides with interior mirrored surfaces, and delivered to interior spaces
via a variety of diffusing devices. In many devices, these diffusers mimic
luminaires. While this can seem counterintuitive – people usually like
the distinctiveness of daylight – it can also be a way to avoid the main
drawbacks of daylight (at least as delivered through conventional ways
such as windows and skylights): glare, e.g., when the orb of the sun is
directly visible to occupants, or thermal discomfort, e.g., when sunlight
falls directly on occupants. Malet-Damour et al. (Malet-Damour et al.,
2020) provide a recent, comprehensive review of the array of options for
different core sunlighting system components.
1.2. Tubular daylight devices
While systems can include solar concentrating dishes and optical
fibers, or mirrored arrays and mirrored ducts, the most common core
sunlighting system is the tubular daylight device (TDD). TDDs are

Abbreviations: DUDI, daily useful daylight illuminance; Eaverage, average horizontal illuminance; Esetpoint, assumed horizontal illuminance setpoint; EUIdaily, esti­
mated daily energy use intensity from electric lighting; EUIt, estimated energy use intensity from electric lighting for timestep t; HDR, high dynamic range; LPD,
lighting power density; MDSA, maximum daily solar altitude; Nall, total number of timesteps used in DUDI calculation; Nuseful, in DUDI calculation, number of
timesteps for which the average horizontal daylight illuminance in the space is in the useful range of between 100 lx and 2 klx; P, lighting system power fraction; Pt,
lighting system power fraction for timestep t; TDD, tubular daylight device.

https://doi.org/10.1016/j.solener.2023.01.022
Received 2 May 2022; Received in revised form 30 December 2022; Accepted 12 January 2023
Available online 27 January 2023
0038-092X/© 2023 International Solar Energy Society. Published by Elsevier Ltd. All rights reserved.
L.L. Fernandes and C.M. Regnier Solar Energy 251 (2023) 420–437

Table 1 producing a livelier, more “daylight-like”, appearance. Carter (Carter,

Overview of independent variables considered in the experiment. 2014) provides a thorough review of the features, physical performance
Category Values and economics of TDDs. More recently, Li et al. (Li et al., 2021) review
the methodologies available for estimating daylight and energy perfor­
TDD configuration dome type clear
prismatic mance of TDDs. Kim and Kim (Kim and Kim, 2010) provide a detailed
diffuser type prismatic review of models for TDD performance analysis and design as well as of
Fresnel different types of systems and several case studies. Malet-Damour
diameter 35 cm (Malet-Damour et al., 2019) provides a detailed review of the prior
53 cm
Solar altitude MDSA low
literature on TDD research.
high While purely theoretical models for calculating TDD performance
Sky cover clear exist in the literature (Laouadi and Atif, 2001; Selkowitz and Johnson,
overcast 1989; Shuxiao et al., 2015), experimental measurements under real sky
partly cloudy
can be more realistic in both the assessment of TDD performance and
providing the basis for mathematical and/or computational models. Of
the experiments reported in the research literature in the last two de­
Table 2 cades or so, several experimental studies were aimed at developing
Overview of dependent variables considered in this experiment. useful correlations and design methods for TDDs. Zhang et al. (Zhang
Category Direct measurements Calculated quantities et al., 2002) develop models for the daylight provision of light pipes
Daylight Horizontal Daily useful daylight illuminance
based on laboratory measurements, performed in Edinburgh, Scotland,
availability illuminance (DUDI) of TDDs of varying diameter. Carter (Carter, 2002) presents experi­
Lighting energy use intensity (EUI) mental results from laboratory and field measurements on light pipes of
Visual comfort Luminance mapping Daylight Glare Probability (DGP) varying diameter conducted in Liverpool, England, and developing
prediction methods for the luminous performance of such devices.
widely available commercially, with products marketed worldwide by Mohelnikova (Mohelnikova, 2009) used measurements of a TDD
large, established companies (Carter, 2014; Malet-Damour et al., 2020). installed in a hallway to validate a theoretical model of tubular light
From top to bottom, TDDs are comprised of a light-transmitting dome, a guides. Lo Verso et al. (Lo Verso et al., 2011) use results from mea­
reflective tube of varying length and shape, and a diffuser. The dome surements and simulations to improve on previous models of TDD per­
admits daylight into the TDD, and it can be made of simple, clear formance. Su et al. (Su et al., 2012) measured the luminous flux output,
polymer, or have some light-redirecting features. The tube is made of a under real skies, of different sizes of commercially-available TDDs and
highly reflective material, in order to minimize light losses as much as developed a model for the luminous flux output of TDDs. Similarly, Patil
possible. The tube can curve and bend in order to accommodate the et al. (Patil et al., 2018) validate existing models and propose new ones
building’s architectural features. This increases flexibility regarding based on experimental measurements of horizontal illuminance in New
where the TDD delivers daylight, as tubes can bring daylight horizon­ Delhi. Malet-Damour et al. (Malet-Damour et al., 2017, 2016) also use
tally or vertical to areas far from the building envelope that would not be measurements under real skies to validate models of the luminous per­
reachable by conventional skylights or windows. Finally, at the point of formance of a TDD. Vasilakopoulou et al. (Vasilakopoulou et al., 2017)
delivery, the tube connects to a diffuser. Diffusers can be chosen with a analyze the performance of a TDD in a test room, using a grid of hori­
variety of optical properties, with some closer to what would be ex­ zontal illuminance sensors, in order to derive correlations between the
pected from a conventional, fluorescent luminaire, and others aimed at magnitude and spatial distribution of indoor illuminance on one side

Fig. 1. Experimental layout.

421
L.L. Fernandes and C.M. Regnier
Fig. 2. View of the test enclosure showing TDD diffuser on the ceiling, pho­
tometers mounted on horizontal rails, and equipment for measuring vi­
sual comfort.
Fig. 3. Photometer on stand inside test cell.
and sky clarity or outdoor illuminance on the other.
Other laboratory evaluations were more generally aimed at investi­
gating the various specific aspects of TDD performance, including
(sometimes in the same study) both commercially available and proto­
type systems. Wu et al. (Wu et al., 2008) investigate two TDD types
under clear sky during the winter in Beijing, comparing two different
types of diffusers, frosted and prismatic, with the frosted diffuser having
slightly improved performance over the prismatic one. Garcia Hansen
et al. (Garcia Hansen et al., 2009) measured the performance of a TDD
with a novel laser-cut dome. Kim and Kim (Kim and Kim, 2010) tested a
TDD with a custom reflector placed inside a clear acrylic dome, and
present results for clear and overcast sky at noon, for a 7x7 grid in a 6 ×
6 × 4 m room in South Korea. In another study (Yun et al., 2010) con­
ducted in the same facility, results for a commercially-available TDD
422
Solar Energy 251 (2023) 420–437
Fig. 4. Glare sensing apparatus.
were shown for various sky types and correlations are developed be­
tween indoor illuminance on one side and sky clarity and solar altitude
on the other. Li et al. (Li et al., 2010) evaluated the light transmission
efficiency and workplane illuminance delivered by TDDs in the hallway
of a commercial building in Hong Kong. Annual potential lighting en­
ergy savings were estimated based on correlations between interior and
exterior illuminance. Baroncini et al. (Baroncini et al., 2010) perform a
1:2 scale test of a novel light pipe concept, including two different types
of diffusers. Detailed results on the differences in performance due to
variation in the type of diffuser are not presented, however. Wu et al.
(Wu et al., 2011) studied the influence of dust and condensation on the
luminous performance of TDDs. Wu and Li (Wu and Li, 2012) measured
the luminous performance of TDDs in two buildings in Beijing, also
computing estimated lighting energy savings. Thakkar (Thakkar, 2013)
measured the illuminance provided by TDDs with varying dome di­
ameters while keeping the tube diameter constant. The effect of varying
the position of a reflector inside the dome has been investigated by Azad
and Rakshit (Azad and Rakshit, 2018) for New Delhi climate. Malet-
Damour et al. (Malet-Damour et al., 2019) show spatial and temporal
illuminance distributions for clear and overcast sky on Reunion Island,
also studying the effects of adding and varying the position of a reflector
inside the dome and of adding a cyclone-resistant subdome under the
main dome. Recently, several researchers have investigated the circa­
dian impacts of the light provided by TDDs (Jain et al., 2019; Malet-
Damour and Fakra, 2021).
On the whole, this literature provides ample evidence on the lumi­
nous and lighting energy benefits of TDDs for an array of component
types, latitudes and climates. However, many of these studies focused on
a single component type (e.g., evaluating a novel dome versus a con­
ventional one, or two types of novel diffusers), were constrained to a few
consecutive days or weeks, and did not cover a representative range of
conditions throughout the year. Additionally, none of them include
measurements of visual comfort quantities. While some studies included
horizontal illuminance measurements on a regular grid, many of them
relied on a reduced number of illuminance sensors to derive horizontal
illuminance trends.
The experiment presented in this paper provides a more compre­
hensive and systematic evaluation of the performance of commercially-
available and commonly-used TDD configuration options than what is
available in the literature to date. To that end, the experiment includes
the main two types of TDD dome that are commercially available:
L.L. Fernandes and C.M. Regnier
Fig. 5. The two types of dome tested: clear (left) and prismatic (right).
Fig. 6. The two types of diffuser tested: prismatic (left) and Fresnel (right). Note that images are underexposed in order to show diffuser detail and aren’t a good
indicator of actual brightness.
Fig. 7. Fourteen-inch frosted round diffuser.
(prismatic and clear), two commercially-available and commonly-used
types of diffuser (prismatic and Fresnel), and two common diameters
for TDDs (53 and 35 cm). These combinations were studied under a
variety of sky types (from completely clear to completely overcast) and a
variety of maximum daily solar altitude (MDSA, or solar altitude at solar
noon) angles. TDD performance variables measured or estimated
included horizontal illuminance on a regular grid, lighting energy use,
and, for the first time in the literature, occupant visual comfort. The
results of this study can inform the research community, as well as de­
signers and potential users of these systems, on what they can expect
from commercially available TDD systems in mid-latitude locations.
423
Solar Energy 251 (2023) 420–437
2. Methodology
2.1. Experimental design
The independent variables considered in this study included envi­
ronmental variables and TDD-specific variables (Table 1). Environ­
mental variables included sky cover – i.e., whether the sky was clear,
partly cloudy or overcast – and MDSA, which varies with the time of
year. TDD characteristics included the dome/tube diameter, the type of
dome, and the type of diffuser.
The dependent variables for this experiment (Table 2) were the daily
useful daylight illuminance (DUDI) (Huo et al., 2020), daily lighting
energy use intensity (EUI), and daylight glare probability (DGP) (Wie­
nold and Christoffersen, 2006). DUDI and EUI were calculated from
measurements of horizontal daylight illuminance in the space lit by the
TDD. DGP was calculated from luminance measurements of the same
space.
Tests were performed under real sky conditions, and for a range of
maximum solar altitudes aimed at spanning, as much as possible, the
range of maximum solar altitude encountered in mid-latitude regions.
The first set of measurements, for low maximum solar altitude, was
performed in February and March 2018 using 53 cm TDDs. In order to
identify whether major differences in performance could be expected
between TDD component types, TDDs from two different major manu­
facturers were used, as well as two different dome and diffuser types. A
narrower set of configurations was used in the subsequent tests with
higher maximum solar altitude, in May and June 2018. These tests also
introduced a new variable – TDD width – with a 35 cm TDD being
L.L. Fernandes and C.M. Regnier Solar Energy 251 (2023) 420–437

Table 3 through the building envelope such as TDDs – in a controlled environ­

Test configurations and dates. ment. This experiment was conducted in one of these cells over the
Configuration Manufacturer Dome Diffuser Diameter Test course of a year.
dates

APP A Prismatic Prismatic 53 cm 9–14 Feb 2.3. Experimental layout

2018
In a 6.27 × 9.60 m test cell, a smaller 4.88 × 4.27 m enclosure was
19–21 created using fire-rated foam boards that went all the way from the floor
May
2018
to the ceiling and were finished with the same paint (0.57 reflectance)
APF A Prismatic Fresnel 53 cm 15–16 that was used in the interior walls of the test cell. The 2.74 m high ceiling
Feb 2018 (0.74 reflectance) is a 61 × 61 cm grid of acoustic tiles, with two LED
luminaires (which were kept powered off for the duration of the
22–23
experiment) and two HVAC diffusers also present within the section of
May
2018 the ceiling that was within the test enclosure. The vertical boards
31 May creating the enclosure were fastened to the metal ceiling grid using
− 3 Jun magnets. The floor is covered with wall-to-wall carpet (0.14 reflectance)
2018 The size of the smaller enclosure was chosen to approximate the largest
4–9 Jan
2019
average floor area per 53 cm TDD in a typical commercial building
ACP A Clear Prismatic 53 cm 17–20 installation, according to available manufacturer literature (Solatube, n.
Feb 2018 d.). The windows in the test cell were blocked using the same type of
ACF A Clear Fresnel 53 cm 22–22 panel used for the interior partitions so that daylight was not measurable
Feb 2018
indoors. Inside the test enclosure, two types of instruments were placed:
26–29 photometers measuring horizontal illuminance and high-dynamic-range
May (HDR) camera apparatuses in order to capture information about glare
2018 from a variety of viewpoints. Fig. 1 shows the experimental layout and
BPP B Prismatic Prismatic 53 cm 24–26 Fig. 2 a view of the test enclosure. More detail on the instrumentation is
Feb 2018
BPF B Prismatic Fresnel 53 cm 27–28
provided below.
Feb 2018
BCP B Clear Prismatic 53 cm 7–9 Mar 2.4. Measurements
2018
BCF B Clear Fresnel 53 cm 10–11
2.4.1. Exterior conditions
Mar
2018 In order to evaluate sky cover, exterior global and diffuse horizontal
irradiance were available from a weather station in a nearby research
9–10 Jun facility, situated approximately 300 m from the test cells.
2018
35A A Prismatic Fresnel 35 cm 5 Jun
2018
2.4.2. Illuminance
Horizontal illuminance inside the space was measured using pho­
11–12 tometers (Licor LI-210R) (LI-COR Biosciences, n.d.), mounted 76 cm (30
Sep 2018 in.) above the floor, and leveled using a bubble level built into the
25–30
photometer base. Fig. 3 shows one of these photometers mounted on a
Jan 2019
35B A Prismatic Frosted 35 cm 7 Jun stand inside one of the test cells. Horizontal illuminance data was used
2018 for assessing light levels within the space.

14–15 2.4.3. Visual comfort

Sep 2018
11–24
Jan 2019
included in the test configurations. In order to capture additional solar
angles with 35 cm TDDs, additional tests were conducted in September
2018 and January 2019, including an additional test with one 53 cm
TDD configuration that did not get sufficient time under clear skies
during the previous low solar angle tests. More detail on the sequence of
the tests and the particular combinations of configurations that were
tested is shown in Section 2.5.
2.2. Facility
The experimental evaluation of TDD performance was conducted at
full-scale in a facility for testing whole-building integrated systems
(FLEXLAB®, Lawrence Berkeley National Laboratory, 2022) in Berkeley,
California, United States. This location, situated at a latitude of 38◦ N,
has a warm-summer Mediterranean climate, with warm, dry summers
and cool, wet winters. The experimental facility has a number of test­
beds, each consisting of two identical test cells, that allow the evaluation
of building technologies and systems – including systems that go
Visual comfort was measured using high-dynamic-range (HDR)
luminance mapping techniques and the DGP metric (Wienold and
Christoffersen, 2006). HDR images were captured using Canon 60D
(Canon USA, n.d.) and Canon 5D (Canon USA, n.d.) digital single-lens
reflex cameras fitted with fisheye lenses and one Licor LI-210 photom­
eter (LI-COR Biosciences, n.d.)(for measuring vertical illuminance),
controlled by a computer running Mac OS (Fig. 4) (Apple Inc, n.d.).
These HDR images were processed in order to calculate a luminance map
of the image, i.e., calculate the luminance for each image pixel.
The DGP metric represents a probability, between 0 and 1, that oc­
cupants of the space will experience glare when their eyes are at the
position of the camera lens at the time that the HDR image was captured.
Subjective ratings corresponding to DGP values are as follows: 0.30,
0.35, 0.40 and 0.45 are the thresholds for “just imperceptible glare” “just
perceptible glare”, “just disturbing glare” and “just intolerable” glare. In
general, it is desirable that DGP remains below 0.35, and that breaches
above that level are of short duration and do not exceed 0.40.
2.5. TDD configurations
The TDDs used in the tests were manufactured by two different major
manufacturers. Each 53 cm TDD was tested with two different domes – a

424
L.L. Fernandes and C.M. Regnier Solar Energy 251 (2023) 420–437

Fig. 8. Occurrence of sky type by TDD configuration for low maximum solar altitude tests.

Fig. 9. Occurrence of sky type by TDD configuration for high maximum solar altitude tests.

Fig. 10. Occurrence of sky type by TDD configuration for medium maximum solar altitude tests.

clear dome and a prismatic dome (Fig. 5) – and two different diffusers at
the bottom – a prismatic diffuser and a Fresnel diffuser (Fig. 6). A
smaller, 35 cm TDD was also tested. It had a prismatic dome; two bottom
diffusers were tested: a 61 × 61 cm Fresnel diffuser similar to its 53 cm
equivalent, and a frosted round diffuser (Fig. 7) 35 cm in diameter. Tests
were conducted between February 2018 and January 2019, in order to
cover a representative range of weather conditions and solar angles. The
configurations tested and the test dates are shown in Table 3.
2.6. Calculations
2.6.1. Sky cover
The type of sky was classified according to the ratio between exterior
direct normal irradiance and diffuse horizontal irradiance, using
thresholds that, in previous research, have been found to be suitable
classifying sky cover and/or controlling automated fenestration systems
under the experiment’s local climate (Fernandes et al., 2013). The sky
was considered clear when this ratio was above 2, overcast when it was
under 0.05, and partly cloudy in between.
2.6.2. Daylight availability
The ability of the TDDs to provide useful interior illumination was
quantified using the daily useful daylight illuminance (DUDI), an
extension, proposed in (Huo et al., 2020), of the useful daylighting
illuminance (UDI) metric (Nabil and Mardaljevic, 2005). It was calcu­
lated by:
Nuseful
DUDI = (1)
Nall
where Nuseful, is the number of timesteps for which the average

425
L.L. Fernandes and C.M. Regnier Solar Energy 251 (2023) 420–437

Fig. 11. Average, maximum and minimum horizontal illuminance obtained Fig. 13. Average, maximum and minimum horizontal illuminance obtained
with 53 cm TDD, prismatic dome and Fresnel diffuser (configuration APF) with 53 cm TDD, prismatic dome and prismatic diffuser (configuration APP)
under clear sky in February (low MDSA). under partly cloudy sky in May (high MDSA).

Table 4
DUDI for three days shown in Fig. 11 to Fig. 13.
Sky Date DUDI (8–18 h)

Clear 15 Feb 87 %
Overcast 01 Jan 0%
Partly cloudy 20 May 100 %

Esetpoint value of 300 lx was used here.1

2. Daily energy use intensity from electric lighting was calculated, for
the hours between 8 AM and 6 PM, by
∑
t=6 PM ∑
t=6 PM
EUI daily = EUI t = Pt LPD (3)
t=8 AM t=8 AM

where EUIdaily is the daily energy use intensity from electric lighting,
EUIt is the energy use intensity from electric lighting for timestep t, Pt
is the lighting system power fraction for timestep t, and LPD is the
installed lighting power density. An LPD value of 9.15 W/m2 was
used here.2
Fig. 12. Average, maximum and minimum horizontal illuminance obtained
with 53 cm TDD, prismatic dome and Fresnel diffuser (configuration APF)
3. Results
under overcast sky in January (low MDSA).

3.1. Sky conditions

horizontal daylight illuminance in the space is in the useful range of
between 100 lx and 2 klx, and Nall is the total number of timesteps. In
this paper, DUDI calculations were calculated for timesteps between 8
AM and 6 PM.
2.6.3. Estimated lighting energy use
The illuminance levels measured at the horizontal workplane were
used to develop estimates of the electric lighting energy use for different
TDD configurations. The method used was as follows:
Daytime (i.e., solar altitude above zero) sky conditions encountered
throughout this study are shown in Fig. 8 to Fig. 10. The sky was clear for
most of the daytime testing time (52 %). The occurrence of partly cloudy
sky was also significant (34 %). The occurrence of overcast sky was
lower but not insignificant (14 %). Based on weather conditions, the test
calendar for the different TDD configurations was continuously adjusted
during the test periods with the goal of achieving at least one full day of
testing with clear sky.Fig. 9.

1. For each timestep, lighting system power fraction P was calculated

as:
⎧
⎪
⎨ 0 if Eaverage > Esetpoint
1
P = Esetpoint − Eaverage (2) This is a commonly used horizontal illuminance setpoint in office spaces
⎪
⎩ if 0 ≤ Eaverage ≤ Esetpoint (David DiLaura et al., 2011).
Esetpoint 2
This value was based on the maximum lighting power allowance that the
2022 version of the California building energy code (California Energy Com­
where Eaverage is the average horizontal illuminance, and Esetpoint is the
mission, 2022) allows for a 300 lx horizontal illuminance level, in a space with
illuminance setpoint assumed for the lighting control system. An the same geometry as the space where measurements were performed.

426
L.L. Fernandes and C.M. Regnier Solar Energy 251 (2023) 420–437

Fig. 14. Illuminance (lx) distribution within interior space obtained with 53 cm TDD, prismatic dome and Fresnel diffuser (configuration APF) under clear sky at
noon during low (left) and high (right) maximum solar altitude tests. Illuminance increased towards the center of the room.

Fig. 15. Illuminance (lx) distribution within interior space obtained with 53 cm TDD, prismatic dome and Fresnel diffuser (configuration APF) under overcast sky at
noon during low maximum altitude test (left). Illuminance (lx) distribution within interior space obtained with 53 cm TDD, prismatic dome and prismatic diffuser
(configuration APP) under partly cloudy sky at noon during low maximum altitude test. Illuminance increased towards the center of the room (right).

3.2. Daylight illuminance
3.2.1. Behavior throughout the day
On a typical clear sky day, TDDs provided a significant amount of
daylight to the interior space, rising in the morning, with a peak around
mid-day and decreasing in the afternoon. For example, on a February
day, a 53 cm TDD with prismatic dome and Fresnel diffuser provided an
average illuminance of at least 200 lx between 9 AM and 4 PM, with
individual illuminance values ranging from around 100 lx to more than
600 lx (Fig. 11). As is to be expected, the amount of light provided when
the sky was overcast was much lower (Fig. 12). Under partly cloudy
skies, performance tended towards intense variability in horizontal
illuminance (Fig. 13). DUDI for the three days shown in Fig. 11 to Fig. 13
is presented in Table 4.
3.2.2. Spatial distribution
In the interior space illuminated by the TDD, horizontal illuminance
at the workplane is consistently higher towards the center of the room.
This is the case under clear sky (Fig. 14). Similar trends were observed
when the sky was overcast or partly cloudy (Fig. 15).
3.2.3. Effect of maximum daily solar altitude
During tests with high MDSA, horizontal illuminance tended to be
higher. This was especially evident in the behavior of maximum illu­
minance, but also in average illuminance as well, even if less markedly
so (Fig. 16). Under clear sky, data from the days shown in Fig. 16 in­
dicates that DUDI is higher for high MDSA tests, but still high for low
MDSA tests (97 % versus 86 %, respectively – see Table 5).
3.2.4. Effect of TDD diameter
Results for 35 cm TDDs showed similar illuminance profiles
throughout the day as obtained with 53 cm TDDs. The main difference in
the results was that illuminance levels were lower than with the larger
diameter devices for both low and high MDSA tests (Fig. 17). When

427
L.L. Fernandes and C.M. Regnier Solar Energy 251 (2023) 420–437

Fig. 16. Average, maximum and minimum horizontal illuminance obtained with 53 cm TDD, prismatic dome and prismatic diffuser (configuration APP) under clear
sky for (left) low and (right) high maximum daily solar altitude tests.

clear domes than for prismatic domes. For high MDSA, trends were not
Table 5
as clear, although higher variability and more profiles with more peaks
DUDI for days shown in Fig. 16.
were observed for clear domes than for prismatic domes (Fig. 19).
MDSA Date DUDI (8–18 h) Average DUDI results are virtually identical between the two types of
Low 13 Feb 86 % domes (98 % and 97 % for prismatic and clear domes, respectively). In
High 21 May 97 % general, results appear to support the assertion that prismatic domes
provide more even illuminance levels throughout the day than clear
domes. Clear domes generally, but not necessarily always, achieve
examining DUDI results, however, it appears that a reduction in TDD
higher average, maximum, and minimum illuminance at some point
diameter has minimal to moderate impact for high MDSA (97 % and 90
during the day than prismatic domes.
% for 53 cm and 35 cm TDDs, respectively), and a significant impact for
low MDSA (86 % and 26 % for 53 cm and 35 cm TDDs, respectively)
(Table 5 and Table 6). 3.2.5.2. Diffuser. Results obtained under clear sky with low maximum
solar altitude did not appear to show any clear trend related to which
3.2.5. Effect of dome and diffuser types diffuser was used, whether regarding the magnitude, shape, or

3.2.5.1. Dome. For days with low MDSA and clear sky, the daily profile
Table 6
tended to be more rounded for prismatic domes than for clear domes
DUDI for days shown in Fig. 17.
both in terms of average and minimum illuminance (Fig. 18). This
translates into somewhat higher average DUDI for prismatic domes MDSA Date DUDI (8–18 h)
when compared to clear domes (86 % versus 80 %, respectively, as Low 25 Jan 26 %
shown in Table 7). Maximum illuminance tended to be more variable for High 05 Jun 90 %

Fig. 17. Average, maximum and minimum horizontal illuminance obtained with 35 cm TDD, prismatic dome and square Fresnel diffuser (configuration 35A) under
clear sky for low and high maximum daily solar altitude tests.

428
L.L. Fernandes and C.M. Regnier Solar Energy 251 (2023) 420–437

Fig. 18. Daily profiles of average, minimum, and maximum illuminance obtained with prismatic and clear domes on low maximum solar altitude days with clear sky.
Note that for one of the clear dome curves (configuration BCP) the sky was not clear until around 1 PM; data is plotted only for the clear sky part of the day.

3.3. Estimated lighting energy use

Table 7
Average DUDI for the same low and high MDSA days shown in Fig. 18 and
3.3.1. Behavior throughout the day
Fig. 19. For clear domes and low MDSA, the incompletely clear day that was only
In terms of lighting energy use, for clear sky days there was a clear
partly plotted in Fig. 18 was not included in the DUDI calculation.
trend of decreasing energy use as the sun rises in the sky, followed by an
Dome MDSA DUDI (8–18 h)
increase as the sun lowers towards the horizon (Fig. 22). During overcast
Prismatic Low 86 % days energy use generally stayed at high levels throughout the day,
Clear Low 80 % especially for low MDSA (Fig. 23). Variability was significant during
Prismatic High 98 %
Clear High 97 %
days with partly cloudy sky (Fig. 24). Daily energy use was clearly
higher for overcast days than for clear days (Table 9).

variability of the daily illuminance profiles (Fig. 20). As a result, sub­
sequent high MDSA tests did not include the full range of dome and
diffuser combinations that were included in the low MDSA tests. Simi­
larly to low MDSA results, for high MDSA no clear trend was observed
(Fig. 21). Average DUDI results (Table 8) appear to indicate minimal
differences in daylight delivery between the two diffusers (97 % and 98
% for high MDSA, 85 % and 83 % for low MDSA, for prismatic and
Fresnel diffusers, respectively).
3.3.2. Effect of maximum solar altitude
For days with clear sky, energy use decreased to zero at some point
during the day, independently of MDSA (Fig. 25). The main difference is
that, for low MDSA, the daily duration of lowest energy use is shorter.
This results in higher daily energy use for low MDSA (Table 10).
3.3.3. Effect of TDD diameter
While still significant and reaching instantaneous levels of less than

Fig. 19. Daily profiles of average, minimum, and maximum illuminance obtained with prismatic and clear domes on high maximum solar altitude days with
clear sky.

429
L.L. Fernandes and C.M. Regnier Solar Energy 251 (2023) 420–437

Fig. 20. Daily profiles of average, minimum, and maximum illuminance obtained with prismatic and Fresnel diffusers on low maximum solar altitude days with clear
sky. Note that for one of the prismatic diffuser curves (configuration BCP) the sky was not clear until around 1 PM; data is plotted only for the clear sky part of
the day.

Fig. 21. Daily profiles of average, minimum, and maximum illuminance obtained with prismatic and Fresnel diffusers on high maximum solar altitude days with
clear sky.

3.3.4. Effect of dome/diffuser type

Table 8
Average DUDI for the same low and high MDSA days shown in Fig. 20 and 3.3.4.1. Dome. When comparing the estimated daily lighting energy
Fig. 21. For prismatic diffusers and low MDSA, the incompletely clear day that use profiles between prismatic and clear domes, similar trends emerge as
was only partly plotted in Fig. 20 was not included in the DUDI calculation.
for illuminance data. When MDSA is low, the profile tends to be slightly
Diffuser MDSA DUDI (8–18 h) wider for prismatic domes and reach zero energy use for shorter periods
Prismatic Low 85 % (Fig. 27). For high MDSA, profile differences between dome are not very
Fresnel Low 83 % evident. In terms of daily energy use, daily EUI results Table 12 indicate
Prismatic High 97 % a consistent advantage of prismatic over clear domes, for both low (34
Fresnel High 98 %
and 38 Wh/m2 for prismatic and clear domes, respectively) and high
MDSA (15 and 17 Wh/m2 for prismatic and clear domes, respectively).
2 W/m2, estimated lighting energy use for the 35 cm TDD was higher
than for 53 cm TDDs (Fig. 26). When comparing daily EUI values be­ 3.3.4.2. Diffuser. Estimated daily lighting energy use profiles do not
tween Table 10 and Table 11 and, reducing TDD diameter approxi­ seem to indicate any clear effect of diffuser type on the daily lighting
mately represents a doubling of daily EUI for low MDSA (34 and 69 Wh/ energy reduction profile (Fig. 28). Daily EUI results (Table 13) suggest a
m2 for 53 and 35 cm TDDs, respectively), and near tripling for high slight but consistent advantage for prismatic over Fresnel diffusers for
MDSA (16 and 45 Wh/m2 for 53 and 35 cm TDDs, respectively). both low (34 and 36 Wh/m2 for prismatic and Fresnel diffusers,
respectively) and high MDSA (16 and 17 Wh/m2 for prismatic and

430
L.L. Fernandes and C.M. Regnier
Fig. 22. Average illuminance and lighting power density obtained with 53 cm
TDD, prismatic dome and Fresnel diffuser (configuration APF) under clear sky
in February.
Fig. 23. Average illuminance and lighting power density obtained with 53 cm
TDD, prismatic dome and Fresnel diffuser (configuration APP) under overcast
sky in January.
Fresnel diffusers, respectively).
3.4. Visual comfort
3.4.1. Behavior throughout the day
Trends were similar to those observed for illuminance levels and
lighting energy reduction. Under clear sky, DGP increases in the morn­
ing and decreases in the afternoon (Fig. 29). In general, DGP levels
peaked in the 0.2–0.3 range, indicating a low probability of visual
discomfort from the TDD diffuser. Viewpoints B, and C tended to have
spend longer periods near their maximum DGP value than viewpoints A
and D. Measured DGP values were insignificant under overcast sky
(Fig. 30); on dark days ambient vertical illuminance levels were some­
times lower than the minimum for triggering the automated DGP
measuring apparatus. As for what was shown earlier regarding hori­
zontal illuminance, DGP variability was clearly higher on partly cloudy
431
Solar Energy 251 (2023) 420–437
Fig. 24. Average illuminance and lighting power density obtained with 53 cm
TDD, prismatic dome and prismatic diffuser (configuration APP) under partly
cloudy sky in May.
Table 9
Daily energy use intensity for days shown in Fig. 22 to Fig. 24.
Sky Date Daily EUI (8–18 h) Wh/m2
Clear 15 Feb 33
Overcast 01 Jan 97
Partly cloudy 20 May 25
days, although without reaching problematic levels (Fig. 31).
3.4.2. Effect of maximum daily solar altitude
Between low and high MDSA tests, the main difference is the dura­
tion of the peak in DGP, around 0.2, with results from high MDSA tests
staying in the vicinity of that maximum value for longer periods, which
is consistent with longer daytime hours (Fig. 32). Another apparent
trend, visible in Fig. 32, is that, for viewpoint A, maximum DGP levels
tended to be observably higher for high MDSA than for low MDSA. For
other viewpoints, results suggest that DGP daily maxima is less affected
by MDSA.
3.4.3. Effect of TDD diameter
When compared with results for 53 cm TDDs, measurements done
with the 35 cm TDDs resulted generally in even lower DGP values,
especially for low MDSA (Fig. 33). With high MDSA, maximum DGP
values reached the vicinity of 0.2 but for shorter duration than with 53
cm TDDs. While there is more of an effect of MDSA on maximum daily
DGP for positions C and D than is observed for 53 cm TDDs, for 35 cm
TDDs the effect for viewpoint A is still the greatest by far.
3.4.4. Effect of dome/diffuser type
3.4.4.1. Dome. For low MDSA tests under clear sky, measured DGP
values were very similar between dome types for viewpoints B and C,
with perhaps the prismatic domes resulting in a smoother, more rounded
daily profile (Fig. 34), i.e., without exhibiting sharp peaks shortly after
sunrise or before sunset. For viewpoint A, results with clear domes
appear slightly higher, peaking as high as 0.24, than with prismatic
domes, which peak at 0.20. In viewpoint D, there was a significant
difference in DGP levels between the two dome types, with the clear
domes reaching peaks about twice as high (around 0.25) as the prismatic
domes. For high MDSA tests (Fig. 35), results were not too different,
L.L. Fernandes and C.M. Regnier Solar Energy 251 (2023) 420–437

Fig. 25. Average illuminance and lighting power density obtained with 53 cm TDD, prismatic dome and prismatic diffuser (configuration APP) under clear sky for
low and high maximum solar angle tests.

• Illuminance increases until solar noon and decreases after that;

Table 10
• Illuminance is highest below the TDD diffuser and decreases with
Daily energy use intensity for days shown in Fig. 25.
horizontal distance from that point;
MDSA Date Daily EUI (8–18 h) Wh/m2 • Useful daylight illuminance increases with MDSA (e.g., for configu­
Low 13 Feb 34 ration APP, DUDI was 97 % on a high MDSA day and 83 % on a low
High 21 May 16 MDSA day);
• Useful daylight illuminance increases with TDD diameter, with the
effect more noticiable when MDSA is low (86 % and 26 % for 53 cm
except for wider peaks for viewpoint A and, for viewpoint D, clear domes
and 35 cm TDDs, respectively) than when MDSA is high (97 % and
resulting in high variability (e.g., swings from about 0.1 to 0.3).
90 % for 53 cm and 35 cm TDDs, respectively);
• Illuminance decreases when going from the center of the room to­
3.4.4.2. Diffuser. No clear differences in visual comfort were observed
wards the periphery;
that could be attributed to the diffuser type, either for low (Fig. 36) or
high MDSA (Fig. 37).

4. Discussion
Table 11
4.1. Daylight illuminance Daily energy use intensity for days shown in Fig. 26.
MDSA Date Daily EUI (8–18 h) Wh/m2
Results for daylight illuminance under clear sky showed the
Low 25 Jan 69
following trends: High 05 Jun 45

Fig. 26. Average illuminance and lighting power density obtained with 35 cm TDD, prismatic dome and square Fresnel diffuser (configuration 35A) under clear sky
for low and high maximum solar angle tests.

432
L.L. Fernandes and C.M. Regnier Solar Energy 251 (2023) 420–437

Fig. 27. Lighting energy use profiles obtained with prismatic and clear domes on clear days for low and high MDSA. Note that for one of the low MDSA curves
(configuration BCP) the sky was not clear until around 1 PM; data is plotted only for the clear sky part of the day.

• Prismatic domes tended to result in a more rounded daily illumi­

Table 12
nance; profile, with higher illuminance in the early morning and late
Average daily energy use intensity for days shown in Fig. 27. For clear domes
afternoon than for clear domes;
and low MDSA, the incompletely clear day that was only partly plotted in Fig. 27
• The two above trends translate into somewhat higher average DUDI
was not included in the EUI calculation.
for prismatic domes when compared to clear domes (86 % versus 80
Dome MDSA Daily EUI (8–18 h) Wh/m2
%).
Prismatic Low 34 • No clear impact of diffuser type was observed; differences in DUDI
Clear Low 38 between diffuser types appear to be minimal.
Prismatic High 15
Clear High 17

• Large (53 cm diameter) TDDs can provide 300 lx average illumi­ Table 13
nance for a significant part of the day (DUDI always in excess of 80 Average daily energy use intensity for days shown in Fig. 28. For prismatic
%). diffusers and low MDSA, the incompletely clear day that was only partly plotted
in Fig. 28 was not included in the EUI calculation.
Additionally, the following features were noticed regarding the type Diffuser MDSA Daily EUI (8–18 h) Wh/m2
of TDD dome and diffuser:
Prismatic Low 34
Fresnel Low 36
• Clear domes tended to result in higher maximum daily illuminance Prismatic High 16
and in a sharper curve; Fresnel High 17

Fig. 28. Daily lighting energy use profiles obtained with prismatic and Fresnel diffusers on clear days for low and high MDSA.

433
L.L. Fernandes and C.M. Regnier Solar Energy 251 (2023) 420–437

Fig. 29. Daylight Glare Probability obtained with 53 cm TDD, prismatic dome Fig. 31. Daylight Glare Probability obtained with 53 cm TDD, prismatic dome
and Fresnel diffuser (configuration APF) under clear sky in February. and prismatic diffuser (configuration APP) under partly cloudy sky in May.

• Differences between diffusers are slight, and results suggest a slight

advantage for prismatic over Fresnel diffusers, for both low (34 and
36 Wh/m2 for prismatic and Fresnel diffusers, respectively) and high
MDSA (16 and 17 Wh/m2 for prismatic and Fresnel diffusers,
respectively).

It should also be noted that the calculation method used in this paper
for daily energy use is aimed at providing a useful general estimate of the
potential differences in performance between different TDD configura­
tions, based on the daylight levels that those configurations are able to
deliver to the workplane. Actual energy use will depend on the partic­
ular specifications of the electric lighting system in use and also on the
ability of a particular lighting control system to take advantage of the
available daylight provided each TDD configuration. As this can vary
significantly between lighting system configurations, the approach
chosen for this experiment was to focus on the ability of TDDs to deliver
daylight, as this is a more intrinsic characteristic of the TDDs themselves
and, therefore, less dependent on the evolution of lighting technologies.

4.3. Visual comfort

Fig. 30. Daylight Glare Probability obtained with 53 cm TDD, prismatic dome
and Fresnel diffuser (configuration APP) under overcast sky in January (note Generally, the DGP levels measured during this experiment were
that there is no data from viewing positions A and D for this day due to the very
consistently below 0.35, indicating a significant probability of visual
low levels of ambient illuminance not being sufficient for triggering the auto­
comfort in spaces where daylight is provided by TDDs. Measured DGP
mated DGP measurements.).
levels tended to drop off with increasing horizontal distance from the
diffuser, although that trend was not strict in the vicinity of the diffuser
4.2. Estimated lighting energy use
(i.e., DGP levels measured from viewpoint A were sometimes lower than
for viewpoints B and C). With the exception of viewpoint A, the effect of
Trends for estimated lighting energy use on clear sky days follow
MDSA on glare did not appear significant. A smaller TDD diameter
those mentioned above for daylight illuminance.
appeared to reduce DGP levels. The effect of dome type did not appear
significant, with the exception of viewpoint D. There was also no clear
• Energy use decreases until solar noon and increases after that;
trend in DGP regarding the diffuser type. This was somewhat surprising
• Energy use increases as MDSA decreases (for APP configuration, 34
because, anecdotally, Fresnel diffusers can produce brighter spots. While
versus 16 Wh/m2 for low and high MDSA, respectively);
this can be considered to add visual liveliness to the interior environ­
• Reducing TDD diameter represents, approximately, a doubling of
ment, one might expect it to cause higher measured DGP whenever the
daily EUI for low MDSA (34 and 69 Wh/m2 for 53 and 35 cm TDDs,
diffuser is in the field of view. It is not clear why this was not observed.
respectively), and near tripling for high MDSA (16 and 45 Wh/m2 for
One possibility is that the images of the bright spots are small enough
53 and 35 cm TDDs, respectively);
that their image is smaller than the pixels of the sensors used to measure
• There is a consistent, if moderate, advantage of prismatic over clear
DGP and therefore their luminance and spatial extent are not correctly
domes, for both low (34 and 38 Wh/m2 for prismatic and clear
captured by such equipment. Further research is needed to better un­
domes, respectively) and high MDSA (15 and 17 Wh/m2 for pris­
derstand this. It should also be noted that, at 0.2 and lower, measured
matic and clear domes, respectively);
differences between DGP values may not correspond to differences in
perceived glare, as DGP’s experimental validation has not focused on

434
L.L. Fernandes and C.M. Regnier Solar Energy 251 (2023) 420–437

Fig. 32. Daylight Glare Probability obtained with 53 cm TDD, prismatic dome and prismatic diffuser (configuration APP) under clear sky for low and high maximum
solar angle tests.

Fig. 33. obtained with 35 cm TDD, prismatic dome and square Fresnel diffuser (configuration 35A) under clear sky for low and high maximum solar angle tests.

Fig. 34. Daylight Glare Probability obtained with prismatic and clear domes on low maximum solar altitude days with clear sky. Note that for one of the clear dome
curves (configuration BCP) the sky was not clear until around 1 PM; data is plotted only for the clear sky part of the day.

435
L.L. Fernandes and C.M. Regnier Solar Energy 251 (2023) 420–437

Fig. 35. Daylight Glare Probability obtained with prismatic and clear domes on high maximum solar altitude days with clear sky.

Fig. 36. Daylight Glare Probability obtained with prismatic and Fresnel diffusers on low maximum solar altitude days with clear sky. Note that for one of the
prismatic diffuser curves (configuration BCP) the sky was not clear until around 1 PM; data is plotted only for the clear sky part of the day.

Fig. 37. Daylight Glare Probability obtained with prismatic and Fresnel diffusers on high maximum solar altitude days with clear sky.

this range.
5. Conclusions
Experimental tests of TDD lighting and visual comfort performance
were conducted for a comprehensive variety of TDD configurations,
including different TDD diameters (53 and 35 cm), dome types
(prismatic and clear), and diffuser types (Fresnel and prismatic). Tests
took place at different times of the year (low and high MDSA) and under
a variety of sky types (clear, overcast, partly cloudy). The results ob­
tained provide quantitative detail about what performance one might
expect from TDDs for a range of TDD configurations and environmental
conditions.
Overall, results indicate that, for clear sky, light levels and potential

436
L.L. Fernandes and C.M. Regnier
lighting energy increase with solar altitude (e.g., for configuration APP,
DUDI/EUI were 97 %/16 Wh/m2 on a high MDSA day and 83 %/34 Wh/
m2 on a low MDSA day) and TDD diameter (e.g., 86 %/34 Wh/m2 and
26 %/69 Wh/m2 for 53 cm and 35 cm TDDs, respectively, on a low
MDSA day; 97 %/16 Wh/m2 and 90 %/45 Wh/m2 for 53 cm and 35 cm
TDDs, respectively, on a high MDSA day). Large (53 cm diameter) TDDs
can provide 300 lx average illuminance for a significant part of the day
(DUDI always in excess of 80 %).
The daily illuminance profile is more rounded for prismatic domes
and has higher peaks for clear domes; this translates into a somewhat
higher average DUDI for prismatic domes when compared to clear
domes (86 % versus 80 %). No clear impact of diffuser type was apparent
in the results.
Measurements indicated no discomfort glare for any of the condi­
tions tested.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
Acknowledgements
The authors wish to acknowledge LBNL colleagues Christian Fitting,
Daniel Fuller, Joshua Mouledoux, and Ari Harding for their invaluable
contributions in setting up and maintaining the experiment; Eleanor Lee
and Christoph Gehbauer for access to solar data.
This work was supported by the California Energy Commission
through its Electric Program Investment Charge (EPIC) Program on
behalf of the citizens of California and by the Assistant Secretary for
Energy Efficiency and Renewable Energy of the U.S. Department of
Energy under Contract No. DE-AC02-05CH11231.
References
Apple Inc, n.d. macOS User Guide [WWW Document]. URL https://support.apple.com/
guide/mac-help/welcome/11.0/mac (accessed 12.7.22).
Azad, A.S., Rakshit, D. (Eds.), 2018. Experimental Study of Tubular Light Pipe System:
Influence of Light Reflector on Its Performance, in: Transition Towards 100%
Renewable Energy: Selected Papers from the World Renewable Energy Congress
WREC 2017, Innovative Renewable Energy. Springer International Publishing,
Cham. https://doi.org/10.1007/978-3-319-69844-1.
Baroncini, C., Boccia, O., Chella, F., Zazzini, P., 2010. Experimental analysis on a 1:2
scale model of the double light pipe, an innovative technological device for daylight
transmission. Solar Energy 84, 296–307. https://doi.org/10.1016/j.
solener.2009.11.011.
California Energy Commission, 2022. 2022 Building Energy Efficiency Standards for
Residential and Nonresidential Buildings.
Canon USA, n.d. Canon Support for Mark II [WWW Document]. URL https://www.usa.
canon.com/support/p/eos-5d-mark-ii (accessed 12.7.22b).
Canon USA, n.d. Canon Support for WWW Document]. URL https://www.usa.canon.
com/support/p/eos-60d (accessed 12.7.22a).
Carter, D., 2002. The measured and predicted performance of passive solar light pipe
systems. Lighting Research & Technology 34, 39–51. https://doi.org/10.1191/
1365782802li029oa.
Carter, D., 2014. LRT Digest 2 Tubular daylight guidance systems. Lighting Research &
Technology 46, 369–387. https://doi.org/10.1177/1477153514526081.
DiLaura, D., Houser, K.W., Mistrick, R.G., Steffy, G.R., 2011. Illuminating Engineering
Society, The Lighting Handbook, Tenth Edition. ed. Illuminating Engineering Society
of North America.
Fernandes, L.L., Lee, E.S., Ward, G., 2013. Lighting energy savings potential of split-pane
electrochromic windows controlled for daylighting with visual comfort. Energy and
Buildings 61, 8–20. https://doi.org/10.1016/j.enbuild.2012.10.057.
FLEXLAB®, Lawrence Berkeley National Laboratory, 2022. FLEXLAB® [WWW
Document]. URL https://flexlab.lbl.gov (accessed 6.13.22).
Garcia Hansen, V., Edmonds, I., Bell, J., 2009. Improving daylighting performance of
mirrored light pipes. Presented at the PLEA2009 - 26th Conference on Passive and
Low Energy Architecture, Quebec City, Canada, p. 6.
Huo, H., Xu, W., Li, A., Cui, G., Wu, Y., Liu, C., 2020. Field comparison test study of
external shading effect on thermal-optical performance of ultralow-energy buildings
in cold regions of China. Building and Environment 180, 106926. https://doi.org/
10.1016/j.buildenv.2020.106926.
437
Solar Energy 251 (2023) 420–437
Jain, S., Fernandes, L., Regnier, C., Garg, V., 2019. Circadian lighting in a space daylit by
a tubular daylight device. IOP Conf. Ser.: Earth Environ. Sci. 238, 012030 https://
doi.org/10.1088/1755-1315/238/1/012030.
Kim, J.T., Kim, G., 2010. Overview and new developments in optical daylighting systems
for building a healthy indoor environment. Building and Environment 45, 256–269.
https://doi.org/10.1016/j.buildenv.2009.08.024.
Laouadi, A., Atif, M.R., 2001. PREDICTION MODELS OF OPTICAL CHARACTERISTICS
FOR DOMED SKYLIGHTS UNDER STANDARD AND REAL SKY CONDITIONS.
Presented at the Seventh International IBPSA Conference. Rio de Janeiro, Brazil,
p. 8.
Li, D.H.W., Tsang, E.K.W., Cheung, K.L., Tam, C.O., 2010. An analysis of light-pipe
system via full-scale measurements. Applied Energy 87, 799–805. https://doi.org/
10.1016/j.apenergy.2009.09.008.
Li, H., Wu, D., Yuan, Y., Zuo, L., 2021. Evaluation methods of the daylight performance
and potential energy saving of tubular daylight guide systems: A review. Indoor and
Built Environment 31 (2), 299–315.
LI-COR Biosciences, n.d. Licor LI-210R Photometric Sensor [WWW Document]. URL
https://www.licor.com/env/products/light/photometric.html (accessed 12.7.22).
Lo Verso, V.R.M., Pellegrino, A., Serra, V., 2011. Light transmission efficiency of daylight
guidance systems: An assessment approach based on simulations and measurements
in a sun/sky simulator. Solar Energy 85, 2789–2801. https://doi.org/10.1016/j.
solener.2011.08.017.
Malet-Damour, B., Boyer, H., Guichard, S., Miranville, F., 2017. Performance Testing of
Light Pipes in Real Weather Conditions for a Confrontation with Hemera. JOCET 5,
73–76. https://doi.org/10.18178/JOCET.2017.5.1.347.
Malet-Damour, B., Bigot, D., Guichard, S., Boyer, H., 2019. Photometrical analysis of
mirrored light pipe: From state-of-the-art on experimental results (1990–2019) to
the proposition of new experimental observations in high solar potential climates.
Solar Energy 193, 637–653. https://doi.org/10.1016/j.solener.2019.09.082.
Malet-Damour, B., Bigot, D., Boyer, H., 2020. Technological Review of Tubular Daylight
Guide System from 1982 to 2020. EJERS 5, 375–386. https://doi.org/10.24018/
ejers.2020.5.3.1809.
Malet-Damour, B., Fakra, D.A.H., 2021. Thermal and spectral impact of building
integrated Mirrored Light Pipe to human circadian rhythms and thermal
environment. International Journal of Sustainable Energy 41 (5), 492–513.
Malet-Damour, B., Guichard, S., Bigot, D., Boyer, H., 2016. Study of tubular daylight
guide systems in buildings: Experimentation, modelling and validation. Energy and
Buildings 129, 308–321. https://doi.org/10.1016/j.enbuild.2016.08.019.
Mohelnikova, J., 2009. Tubular light guide evaluation. Building and Environment 44,
2193–2200. https://doi.org/10.1016/j.buildenv.2009.03.015.
Nabil, A., Mardaljevic, J., 2005. Useful daylight illuminance: a new paradigm for
assessing daylight in buildings. Lighting Research & Technology 37, 41–57. https://
doi.org/10.1191/1365782805li128oa.
Patil, K.N., Kaushik, S.C., Garg, S.N., 2018. Performance Prediction and Assessment of
Energy Conservation Potential for a Light Pipe System in Indian Composite Climate
of New Delhi. Journal of Solar Energy Engineering 140, 051012. https://doi.org/
10.1115/1.4039656.
Selkowitz, S., Johnson, K., 1989. Light Guide Design Principles (No. LBL-20546).
Lawrence Berkeley National Laboratory.
Shuxiao, W., Jianping, Z., Lixiong, W., 2015. Research on Energy Saving Analysis of
Tubular Daylight Devices. Energy Procedia 78, 1781–1786. https://doi.org/
10.1016/j.egypro.2015.11.305.
Solatube, n.d. Solatube daylighting systems spacing criteria.
Su, Y., Khan, N., Riffat, S.B., Gareth, O., 2012. Comparative monitoring and data
regression of various sized commercial lightpipes. Energy and Buildings 50,
308–314. https://doi.org/10.1016/j.enbuild.2012.03.053.
Thakkar, V., 2013. Experimental study of Tubular Skylight and comparison with
Artificial Lighting of standard ratings. International Journal of Enhanced Research in
Science Technology & Engineering 2, 6.
Vasilakopoulou, K., Kolokotsa, D., Santamouris, M., Kousis, I., Asproulias, H.,
Giannarakis, I., 2017. Analysis of the experimental performance of light pipes.
Energy and Buildings 151, 242–249. https://doi.org/10.1016/j.
enbuild.2017.06.061.
Wienold, J., Christoffersen, J., 2006. Evaluation methods and development of a new
glare prediction model for daylight environments with the use of CCD cameras.
Energy and Buildings 38, 743–757. https://doi.org/10.1016/j.enbuild.2006.03.017.
Wu, Y., Jin, R., Li, D., Zhang, W., Ma, C., 2008. Experimental investigation of top lighting
and side lighting solar light pipes under sunny conditions in winter in Beijing, in:
Wang, A., Liao, Y., Song, A., Ishii, Y., Fan, X. (Eds.), . Presented at the International
Conference of Optical Instrument and Technology, Beijing, China, p. 71571O.
https://doi.org/10.1117/12.811992.
Wu, Y.P., Li, J., 2012. Analysis of Energy Saving Effect of Solar Light Pipe Systems in
Beijing Olympic Buildings. AMR 452–453, 294–298. https://doi.org/10.4028/www.
scientific.net/AMR.452-453.294.
Wu, Y.P., Wang, X.D., Chen, Z.G., Zhang, C.Y., 2011. Experimental Study on the
Influence of Daylighting Performance of Solar Light Pipes by Dusts and
Condensation. AMR 374–377, 1096–1099. https://doi.org/10.4028/www.scientific.
net/AMR.374-377.1096.
Yun, G.Y., Shin, H.Y., Kim, J.T., 2010. Monitoring and Evaluation of a Light-pipe System
used in Korea. Indoor and Built Environment 19, 129–136. https://doi.org/10.1177/
1420326X09358007.
Zhang, X., Muneer, T., Kubie, J., 2002. A design guide for performance assessment of
solar light-pipes. Lighting Research & Technology 34, 149–168. https://doi.org/
10.1191/1365782802li041oa.